Impact of COVID-19: A Text Mining Analysis of Twitter Data in Spanish Language

Abstract

Background:

Latino communities in the United States and Latin America are disproportionately affected by the COVID-19 pandemic. We analyzed information shared on Twitter in Spanish language for insights into the public’s communication and information needs about the COVID-19 pandemic.

Methods:

We performed a mixed-methods analysis using a text mining approach. We used SAS Text Miner, an algorithmic-driven statistical program to capture 10,000 tweets posted between June 3, 2020, and June 10, 2020. We used the following search terms to capture relevant Twitter messages in Spanish language: “coronavirus,” “covid-19,” “corona,” and the hash tags “#COVID19” and “#Coronavirus.” Key text topics were identified and categorized into themes using an emergent content analysis.

Results:

We identified 12 text topics and six themes: (1) prevention measures, (2) epidemiology/surveillance, (3) economic impact, (4) optimizing nursing workforce, (5) access to reliable information, and (6) call for a response from the local government. Top trending hashtags from our search included #COVID19 (n = 7,098), #Coronavirus (n = 6,394), and #SNTESALUD (n = 2,598).

Conclusions:

Spanish-language Tweets related to the COVID-19 pandemic contained information from health departments and labor unions on the surveillance, prevention, and impact of COVID-19. Public health officials should consider increasing their use of Twitter to ensure a wide dissemination of messages about COVID-19 in Spanish outlets.